In the former part of tbis paper the author had pointed out various methods of producing coloured rings between surfaces in contact, and had proved that no other surfaces are concerned in their formation; and he now proceeds to show, that, by varying the figures of the surfaces, at corresponding change of the form of the rings will be ob- served. As a spherical surface applied to a plane surface produces circular rings, so, when it is applied to a cylindrical surface, the rings become ellipses ; and when the cylinder is applied to a plane, the ellipses become straight lines, and irregular curves occasion irre‘ gularity in the form of the rings. But Dr. Herschel is of opinion, that plane surfaces, applied to each other at any extremely small angle, cannot exhibit any such appearances of colour unless they are dis- figured by unequal pressure. .
The author next examines the circumstances of the prismatic blue bow observed by Newton, as the limit to the perfect reflection that occurs in the interior of a prism when light is much inclined to its surface; and Dr. Herschel calls this the critical separation of the colours. He also examines the prismatic red bow observable at the same critical inclination, when the transmitted light prevails over that which is reflected; and remarks the change from red to blue, occa— sioned by changing the direction of the light.
The next phenomena observed are the coloured streaks that are produced adjacent to these bows when the reflecting surface of the prism is in contact with another surface; the streaks being straight when the surfaces are plain, or curved, according to their different kind or degree of curvature. In the experiments .which follow, this efi'ect of a surface in contact with the prism is employed as a crite- rion for distinguishing more clearly the acting surface. I
Various more complicated appearances are next examined; in which, by reason of second reflection, the origin of the bows is less distinct, or in which two bows may be seen at once from the parti- cular form of the prism.
In the formation of bows, it is shown, that only one surface of the prism is concerned by roughening other surfacw, and by applying to them various glasses, without preventing the appearance of the bows.
Since the streaks which are seen beneath the blue how, when a plane surface is in contact with a prism, contain all the prismatic colours, it is evident that the less refrangible, after having been transmitted, are reflected back by the contiguous surface, and then re-enter the prism.
Dr. Herschel next proceeds to delineate, from strict computation, the course of the several rays; and, by a diagram drawn upon an enlarged scale, shows, that when light falls upon a pair of surfaces nearly in contact, and when single rays or small pencils are separated
